International Journal of Information Science and Technology

Vol 5, No 2, ISSN 1726-8125, 2007

DOES SIMPLE BINARY CROSSOVER HASTEN RCGA CONVERGENCE?

C. Lucas, Ph.D. , Center of Excellence on Control and Intelligent Processing, ECE Dept., University of Tehran, Iran; and SIS, IPM

A. Nayyeri, M.S. , Faculty of Electrical and Computer Engineering, University of Tehran, I. R. of Iran

M. Yabandeh, M.S. , Faculty of Electrical and Computer Engineering, University of Tehran, I. R. of Iran Corresponding Author

Abstract

Real Coded Genetic Algorithm, RCGA, is the type of GA which operates on chromosomes with real valued parameters. Different mutation and crossover operations are defined for RCGA. One usable crossover for this kind of GA is to consider its chromosomes simply as bit strings and utilize the same operations as Binary Coded GA. In this paper, we attempt to show that this kind of crossover can not hasten the convergence process unless the break points fall at the boundaries of parameters in the chromosome.

Keywords - BCGA, Convergence Time, Crossover, Genetic Algorithm, RCGA, Real Numbers.

email: lucas@ipm.ir

email: a.nayyeri@ece.ut.ac.ir

email: m.yabandeh@ece.ut.ac.ir